Interesting fuel issue

[Nev]

Yesterday my friend Jason and I did the fuel flow testing for my Bearhawk.

We checked flow rates with the electric fuel pump on and off, fuel caps on and off, tail low, cruise attitude, nose very low, etc, and any combination we could think of. We also ran a tank very low, unported the front fuel line, turned the pump on, and tried to suck air. However it kept pumping fuel right to the end.

One issue had us scratching our heads. When a tank was completely empty, if we added 5-10 liters, no fuel would come out. If we then selected another tank, or BOTH, or turned the pump on, then selected back to the original tank, the problem would resolve immediately back to normal flow. We repeated this scenario at least 10 times, and on both L&R tanks. Initially we thought we had a blockage, but it became apparent that we were dealing with an air lock in the section of tubing between the electric pump and the gold cube transducer. On several occasions we left the fuel Selector in with the L or R position while pouring fuel in the tank, and none came out. However anytime that there was already a small amount of fuel in the tank, flow was completely normal - as the line was primed.

It is very unlikely to be of any consequence because it's highly unlikely that a tank will ever start completely empty, and only 5-10 liters with be added. Even then, with BOTH selected, the problem didn't occur or it self cleared We think that when first adding fuel to a dry tank an air lock is forming until enough head pressure is present to overcome it, or additional pressure or venting is present.

It may be specific to my installation. However, once a flow was established, it always flowed again even after multiple fuel Selector positions.

I decided to add this post in case anyone else has a similar head scratching when performing fuel flow checks, or in case this scenario triggers a light bulb moment that something more serious might be occurring, or maybe has the potential to occur in our collective fuel systems. Given that it occurred with the electric pump turned OFF, I suspect it's probably not confined just to a system with a fuel pump.

[AKKen07]

Really interesting Nev, thanks for sharing. Will have to look for this in my installation as well since it is very similar.

[Nev]

Giving it some more thought, it could be of consequence if a tank is low, and becomes unported. This could potentially completely clear a line of fuel, including the gascolator. I think in that scenario the line may be susceptible to an air lock. An example would be side-slipping the aircraft with say 1/4 tanks remaining ( reasonably common).

If the line does air lock (indicated by the engine stopping), it appears that changing tanks to any other tank would resolve it. Also turning the fuel pump on resolved it for us. Presumably if a pump is fitted and a side-slip manoeuvre is being performed then the pump is most likely already on. However having the pump on (or an engine driven pump) may also give a false sense of security, because if a tank remains unported then the pump may suck air (not possible in a gravity feed situation). But I think this should resolve once the port becomes flooded again.

A Header tank may go some way to offering protection against the air lock issue, but I think it would need to be tested very very thoroughly. Because simply installing a header tank might mask an underlying issue. The fuel in the header tank might appear to cover short term fuel interruptions, but if a line is airlocked, the header might become completely empty and take much longer to get the engine running again. Also, having a header tank doesn't guarantee that an airlock won't occur between the header and the engine where resources to resolve it might be limited.

I'm not an authority on any of this and my knowledge is limited, so don't take any of it as gospel, just thoughts based on what we saw.

[AKKen07]

Have you changed anything as a result or do you feel comfortable with your system as is?

[Sir Newton]

May I ask, what is the line size that is providing static head pressure from the tank? What type of pump have you installed?

[Hewko]

Nev,
this same situation with gravity flow systems has been documented by some builder's of Murphy Rebel's, so it's not an uncommon issue. Going to 'Both' allows the air lock to clear itself. I think those guys put a tank switch procedure in their operating manual to deal with it. I haven't experienced it myself, even doing slips with low fuel, on both tanks, but it's something to be aware of.

[Mark Goldberg]

I had exactly the same thing occur in my 4 place when we were first putting fuel in it. At that time I only had a L/R fuel selector valve. So trying the both setting was not an option to try. I found that it took 10 gallons or so in a tank to push the air put of the way to start to fuel flowing. At the time we attributed the problem to my fuel lines from the tanks into the fuselage going very slightly UPHILL.

Still, it would have seemed a little head pressure from a few gallons would have started the fuel flowing into the completely empty fuel lines. But it took much more fuel than expected.

At the time I knew a man who was the father of my son's friend. He is a PHD Physics. I asked him how it was possible what we saw to occur. He responded that sometimes fluid does behave as expected. And theoretically it should have flowed with just a little head pressure. But the air block in essence was holding on for much longer than we might have thought.

So the lesson for me then was to not completely run a tank dry and then just add 5 gallons. Get at least 1/2 full to assure the fuel would flow afterward. After 1300 hours it has never been a problem. Mark

[AKKen07]

Great info Mark

[Nev]

Have you changed anything as a result or do you feel comfortable with your system as is ?

Ken - I'm happy with the system now and won't be changing anything. My main concern was to make sure there wasn't a blockage from swarf etc.

Also, we made sure that the air lock could be cleared with a simple memory action. The initial Engine Failure procedures etched firmly in my mind from the past 35 years are "Fuel pump on, change tanks". This clears the issue, no matter whether you select L, R, or BOTH. Incidentally, if the pump is already on, once I've tried both tanks and a problem still exists I would probably turn the pump off and select the lowest tank. But that's in case of a vent blockage and is a different issue.

Sir Newton - I've got the standard 3/8" fuel line, and an EFII fuel pump. I believe the pump itself is actually a Walbro pump that is commonly used in cars.

[Sir Newton]

Understood, that is a very good pump. Out of curiosity does the tank have a 1/4 or 3/8 ntp welded connection for engine supply?
There was a comment previously in the thread that mentions the routing of the tubing going up hill. When speaking about gravity fed systems there are known factors for every thing component that is in the system. ex. 45deg vs 90deg bends & so on. So everything in this system can be calculated with some effort. Do not run 3/8 tubing up hill period

[Nev]

Yeh - the transducer instructions had me running a short section uphill to and from the transducer. Inflight this may be level perhaps.

[Sir Newton]

I am concerned about this. Thus been doing some calculations on the system I am designing for my western engine install.
Back of the napkin notes are telling me in tank pumps may be the safest option. In my situation nothing is final & every option is on the table to be explored.

[svyolo]

Both tank outlets are close to the bottom, but not quite at the bottom of the tanks. I think if the tanks are draining completely dry, the last gallon or two is probably not being gravity fed per se, but maybe actually siphoning out, which technically is also caused by gravity. But it is not flowing straight downhill. It would have to be sucked up and over the little hump by weight of the fuel in the vertical lines. Without fuel in the lines, the siphoning stops.

Just a theory.

[Nev]

Both tank outlets are close to the bottom, but not quite at the bottom of the tanks. I think if the tanks are draining completely dry, the last gallon or two is probably not being gravity fed per se, but maybe actually siphoning out, which technically is also caused by gravity. But it is not flowing straight downhill. It would have to be sucked up and over the little hump by weight of the fuel in the vertical lines. Without fuel in the lines, the siphoning stops.

Just a theory.

We could see the fuel level in the sight gauge, so fairly sure it wasn't relying on siphoning.

[svyolo]

Yeah, just thinking out loud. On my tanks, the lower sight gauge port is lower in the tank/closer to the bottom, at least in the level flight attitude. Just a much smaller weld on bung, the bottom of which can sit much closer to the bottom of the tank.

[davzLSA]

Nev, I'm just curious, did you ever try your tests with the flow transducer removed from the system?

[Nev]

Dave, I didn't try this. Although today I fixed a sight glass leak and then ran some fuel through the system and out through the gascolator drain valve (removed it). I noticed that there was no air lock.

[Gerhard Rieger]

Neville, you mentioned a header tank, I researched the fuel system used in the Bearhawk especially the un-porting of a fuel line. My conclusion was that if you use a fuel injected engin, specifically the IO540 Lycoming, a header tank will be needed to provide uninterrupted fuel flow in all attitudes of flight.
I know there are many opinions on this matter and it needs to be planed carefully. Here is what I came up with.

Blue: fuel from L&R main tanks to fuel selector, 3/8’
Red : fuel selector to gascolator 3/8’
White : Gascolator to header tank 3/8’
Green : Header tank to fuel pump 3/8’
Green/Red : fuel pump to engin 3/8’ ( no return to the header tank )

Yellow : Two 1/2’ vent lines returning to the main fuel tanks highest point. One vent is in the right front corner and the other is in the rear left corner. This way you will always have a vent in a high point no matter what attitude you have the aircraft in. The Header tank has a 2,2 Gal capacity and it will remain full at all flight attitudes. Having the header tank set up likes this gives you the assurance that fuel will easily travel from the mains through the 3/8 lines, fuel selector, gascolator,( all constrictors of flow) because of the large vent lines at their end destination. The electric fuel pump only have to take the fuel a short distance to the engin, once you have that pressurised the manual engin driven fuel pump will have no problem to get the fuel the engin needs at all times. With this setup you don’t even need to put the fuel pump on when changing tanks. THERE IS ONE RISK WITH THIS SYSTEM, THAT IS NOT OPENING THE FUEL SELECTOR VALVE DURING YOUR PREFLIGHT AND STARTUP PROCEDURES . If standard checklist and operating procedures are followed like we all have been taught and use, it will never be a problem, still, I am planing a safety pin through the fuel selector arm to the throttle to eliminate this risk. Hope this is helpful to someone, Have fun testing your awesome build Bearhawk.

[robcaldwell]

The fuel injection system on my IO-540 is from AirFlow Performance. I personally met with and spent an afternoon with Don Rivera, the owner of the company. I also returned and spent 2 days attending his Fuel Injection 101 course. I won't go into the details here, but Don demonstrates why his system does not require a header tank. This might be different on other systems such as Bendix.

[Gerhard Rieger]

Hi Rob, I am using the Bendix system, like I said there are many variants , but this is what have done. Missing seeing more video of your BH

[Battson]

I image that a little suction from the engine or electric pump would solve this, as you said it did. Practically, once there's an engine connected to the fuel system *unless you're gravity feed, the problem cannot exist.

[JimParker256]

One thought on header tanks: It is super-easy to add a low-fuel warning light to the header tank, so that you will know that fuel is no longer refilling the header. I did this on my RANS S-6 when we installed a 2.75 gallon header tank. The low-fuel light comes on when the level drops below 2 gallons. That will hopefully give me enough time to get it on the ground safely before fuel starvation kicks in.

For that low-fuel warning light, a good (cheap) option is the Radiant Instruments Bingo 4 available through Aircraft Spruce. It doesn't require a hole in the tank, since it uses ultrasonic technology to detect the liquid level. I like the version with the panel lights (green = good, red = low level).